Game UI Voice Pack converts approved game interface copy into 8 to 20 individually labeled, reviewable voiceover clips.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install game-ui-voice-pack
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install game-ui-voice-pack using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Game UI Voice Pack?
Game UI Voice Pack is an Openclaw Skills workflow for producing one audio file per labeled game interface cue, including buttons, confirmations, warnings, wins, failures, combos, and level-up messages. It uses only the written UI lines supplied by the studio and prioritizes consistent voice, pronunciation accuracy, and clear clip-level delivery.
The skill supports catalog voices and authorized cloned voices through Beatra speech tools. It separates planning, clone approval, speech generation, billing confirmation, task polling, audio review, and delivery so studios can stop safely before any paid operation. Openclaw Skills users receive a transparent slot list before generation and actionable recovery behavior when responses or task identifiers are lost.
Game UI Voice Pack Use Cases
- Create a complete game UI voice pack from existing button, win, fail, and status copy.
- Produce a default 10-clip set covering click, confirm, cancel, start, pause, win, fail, combo, warning, and level-up cues.
- Generate 8 to 20 consistent clips using a catalog voice or an authorized cloned voice.
- Prepare labeled interface audio for prototypes, live-service games, mobile games, and accessibility-focused game builds.
- Verify names and coined terms against a studio pronunciation table before audio generation.
- Estimate live speech or voice-clone costs and review available wallet credits before submitting paid work.
- Recover safely when a create response is lost by searching and verifying existing tasks instead of replaying requests.
- Use Openclaw Skills when the studio needs UI voice clips rather than scripted character dialogue or generic narration.
How Game UI Voice Pack Works
- Collect the hard inputs: approved UI lines, pronunciation guidance for names or coined words, desired pack size, and voice or likeness rights when cloning is requested.
- Validate scope. Route character dialogue to the dedicated game-script voice workflow and generic narration to a narration workflow; this skill reads only supplied UI copy.
- Build a free labeled slot list, defaulting to 10 cues or using a confirmed count between 8 and 20. Record each spoken line and whether it uses a catalog voice or clone.
- Select safe defaults: one speech call per slot, MP3 output, speed 1.0, and one brand voice across the pack. Keep each submitted input at or below 50,000 characters.
- If cloning is requested, inspect an authorized voice sample, read the live voice-clone pricing card, obtain confirmation, upload local audio only through the bundled client, and preserve the returned artifact ID.
- Before speech generation, call
beatra.models.listfor thetext_to_speechcapability and usebeatra.voices.listwhen a catalog voice still needs to be selected. - Show a separate six-field speech card containing work, live credits, call count, opaque request identity, stop behavior, and insufficient-balance handling. Wait for approval before paid speech.
- Submit exactly one
beatra.speech.synthesizerequest per named slot throughscripts/mcp_client.py, using a new opaqueclient_request_idfor every slot. - Poll each returned task with
beatra.tasks.get, then inspect actual audio MIME type, duration, file size, andbilling.net_charged_credits. - Review every clip for consistent voice, correct pronunciation, and exact alignment with the written UI list. If a response is lost, use
beatra.tasks.listand verify withbeatra.tasks.getbefore considering replay. - Deliver the labeled audio files and report only qualities the host can actually hear. A changed line, voice, or speed requires a new request card and new request ID.
Game UI Voice Pack Setup
- Install or open the skill package containing the bundled
scripts/mcp_client.pyclient and its referenced documentation. - Complete the non-billable registration and authentication process described in
references/installation-and-auth.mdandreferences/installation-registration.md. - Invoke all remote Beatra operations through the bundled client. Do not configure a host Beatra Connector and do not use REST or OpenAPI as a fallback.
python3 scripts/mcp_client.py call beatra.models.list
{"capability":"text_to_speech"}
- Prepare the approved UI copy, pronunciation table, target clip count, and voice-rights evidence before planning or submitting work. Missing pronunciation readings for names are a hard stop.
- For an authorized local clone sample, inspect it first and upload it only through the bundled client using
beatra.assets.upload; never pass a local filesystem path directly to voice cloning or speech synthesis. - Keep the installation current with its built-in automatic update controls. Updates check at most once every 24 hours per installation and verify package-owned files before replacement.
python3 scripts/mcp_client.py update --auto off
python3 scripts/mcp_client.py update --auto on
python3 scripts/mcp_client.py update --check
- For account questions, use
beatra.wallet.getfor remaining credits or live estimates andbeatra.wallet.ledgerfor charges. Do not make a wallet check mandatory before every paid request. - Read the referenced workflow, task-recovery, billing, MCP diagnostics, update-safety, and uninstall documentation before operating a production Openclaw Skills voice pipeline.
Game UI Voice Pack Data Schema & Taxonomy
Core inputs
| Field | Purpose |
|---|---|
| UI copy list | The only approved source for spoken lines |
| Pronunciation table | Required readings for names and coined terms |
| Pack count | Number of clips, normally 8 to 20; default is 10 |
| Voice mode | Catalog voice or authorized cloned voice |
| Voice rights | Consent and likeness authorization required for cloning |
voice_id |
Frozen voice identifier used consistently across the pack |
Slot taxonomy
The default labeled slots are click, confirm, cancel, start, pause, win, fail, combo, warning, and level-up. Each slot stores its label, the exact written UI line, and the selected voice mode. The free slot list remains usable even if the seller stops before paid generation.
Request and task metadata
- Clone operations use one six-field approval card, one live
voice_cloneprice, one authorized sample, and one opaqueclient_request_id. - Speech operations use one six-field approval card per slot, one live
text_to_speechprice multiplied by slot count, and one new opaqueclient_request_idper slot. - Local samples are represented by the artifact ID returned from
beatra.assets.upload, never by a local path. - Returned
task_idvalues are polled throughbeatra.tasks.get; missing create responses are recovered throughbeatra.tasks.listfollowed by verification. - Reuse of an ID is allowed only for byte-identical arguments. Changed copy, voice, or speed creates a new request.
Generated deliverables
Each completed slot produces one labeled audio clip, normally in MP3 format at speed 1.0. Delivery metadata should include the actual audio MIME type, duration, file size, task result, and billing.net_charged_credits. Estimates are not treated as final charges. Review status records voice consistency, pronunciation alignment, and whether the clip matches the approved UI list.
Game UI Voice Pack Advanced Features
- Separate clone and speech stages with independent approval cards, billing checks, counts, and request identities.
- Live pricing discovery through model and voice capability cards instead of remembered credit values.
- Catalog voice discovery with
beatra.voices.listand frozenvoice_idusage for pack consistency. - Authorized voice-sample upload through
beatra.assets.upload, with artifact ID preservation and local-path isolation. - Idempotent task recovery using
beatra.tasks.listandbeatra.tasks.getbefore replaying any request. - Per-slot synthesis architecture that keeps each button, win, fail, or warning cue independently labeled and replaceable.
- Strict input controls that cap each submitted speech input at 50,000 characters and encourage sentence-boundary splitting.
- Exact insufficient-balance handling that relays the official message and top-up URL without retrying until the seller confirms funding.
- Read-only wallet inspection through
beatra.wallet.getandbeatra.wallet.ledger. - Verified automatic updates from fixed official discovery and immutable CDN paths, with rollback and fail-open recovery that preserves the current installation.
- Safety boundaries that prohibit invented lines, unverified consent, host connector calls, REST fallbacks, and duplicate paid submissions.
- A practical Openclaw Skills workflow that distinguishes UI voice production from character dialogue and general narration.
SKILL.md
Loading
Related Openclaw Skills
Open Claude
Open Claude launches Claude AI directly in the Brave browser so you can begin working with Anthropic’s assistant immediately.
Open Grok
Open Grok launches Grok AI at grok.x.ai in Brave so you can start chatting with xAI using a simple command.
Open Claude
Open Claude launches Claude AI at claude.ai in Brave so you can start an Anthropic conversation instantly.
Skill Injection Scanner
Skill Injection Scanner audits agent skill files for hidden instructions and prompt-injection patterns before they can manipulate your AI agent.
CoinGecko Crypto Price Skill
A secure, read-only crypto market data skill that fetches live prices from CoinGecko with automatic Kraken fallback.
Enterprise GEO Optimization
An end-to-end enterprise GEO operating system for diagnosing, improving, distributing, measuring, and iterating brand visibility in generative search.
Featured*
